Big Database Research Articles

MODCN: Fine-Tuned Deep Convolutional Neural Network with GAN Deployed to Forecast Diabetic Eye Damage in Fundus Retinal Images

Diabetic Retinopathy (DR) and Glaucoma are two of the most common causes of vision loss world-wide. However, it can be averted if therapy is begun early enough. In biomedical applications, the use of digital image processing has assisted in the automated identification of some ailments at an earlier stage. To make this prediction generally neural network classifier models were previously used, but these models have the drawback of being unable to detect multiple illnesses that occur in the eye at the same time and require a big database for successful classifier training. As a result, a model is needed to reliably distinguish DR and Glaucoma in diabetic individuals more accurately and with minimum dataset images. In this view, this study introduced Mayfly Optimized Deep Convolutional Network (MODCN) model for automated disease detection in the fundus retina images. In the MODCN model, the images are initially preprocessed, segmented at generator in the GAN model then a discriminator readily gives synthesis of real images of the fundus retina, thus a wide database has been created and considered as training images for the MODCN classifier. MODCN classifier has a modified high-density layer as a transition layer to avoid overfitting and the errors are minimized by tuning the hyperparameters using Mayfly Optimization Algorithm. After feature mapping, the classes normal, DR and Glaucoma are labeled and stored. At the testing stage, images are preprocessed, feature mapped and classified in the MODCN model. Thus, the proposed MODCN model detects multiple illness such as Diabetic Retinopathy and Glaucoma at the same time even with a small amount of database that performs a successful classifier training. This model is then evaluated and gives an accuracy of 99% that was higher compared to previous models.

Data Fusion Approach to Identify Distribution Chain Segments in Freight Shipment Databases

Understanding the logistic determinants of freight trips is an important goal in freight transport modeling. Freight shipments move between nodes in the supply chain for different logistic purposes, including production, storage, transshipment, and consumption. A key problem with data availability is that databases generally do not identify these purposes, given the commercial sensitivity of the data. In addition, including information on senders and receivers of the shipments is often prohibitively costly. Therefore, one of the challenges of transport data analysis is to identify freight trip purposes using data fusion, linking information about the main function of logistics nodes to trips in existing databases. This paper proposes a data fusion approach to enrich big truck shipment databases with firm registry data. We use the national freight shipment micro-database from the Netherlands which includes shipment, vehicle, and tour information. Although our presentation here uses formats and methods of accounting for freight data used in the Netherlands, it can be readily replicated for conditions in other countries, as long as similar data sets on shipment data and firm registry are available. The enriched, new database contains transport and firm data for more than 2 million observed trips with information on the vehicle used, shipments carried, and sender/receiver firm. An initial descriptive analysis provides unique empirical insights into the logistic determinants of freight trips. These include the share of national trips that use intermediate nodes, typical changes in shipment sizes, and the role of distribution centers for (de)consolidation of shipments.

Use of blockchain technology in veterinary practice

The article discusses the use of blockchains in various areas of veterinary medicine, including veterinary education. So, in the Russian Federation, as early as 2019, a departmental project "Digital Agriculture" was created, which includes three stages: the creation and integration of databases, the introduction of the "Agricultural Solutions" module of the national platform, and the final one - the creation of a system for continuous training of specialists of agricultural enterprises in order to form of them competencies in the field of digital economy.The polythematic abstract-bibliographic base is analyzed. The main attention was paid to such scientometric databases as: Web of Science Core Collection, Medline, PubMed, RSCI, as well as data from the eLIBRARY.ru portal. An assessment of the statistical data and publication activity of the authors regarding the request and use of blockchain technology was made.The aim of the study was to analyze the publication activity of the authors and relevant data search from 2019-2023, regarding the use of blockchain technologies in veterinary medicine.If in 2019, at the request of "blockchain in veterinary medicine", the system issued 7 publications, for 2020 - 15 articles, 2021 - 16 articles, then in 2022. already 25 articles. As of November 2022, in 2023, there are already 4 preprints. All these data indicate that this technology is rapidly gaining momentum in the scientific field, and many scientists and entrepreneurs see great prospects in it.According to the studies conducted in scientometric databases, and their distribution by relevance, showed the relevance of introducing this technology, big database technology, blockchain technologies not only in veterinary practice, but also in veterinary education.

Data-driven short circuit resistance estimation in battery safety issues

Developing precise and fast methods for short circuit detection is crucial for preventing or mitigating the risk of safety issues of lithium-ion batteries (LIBs). In this paper, we developed a Convolutional Neural Networks (CNN) based model that can quickly and precisely predict the short circuit resistance of LIB cells during various working conditions. Cycling tests of cells with an external short circuit (ESC) are produced to obtain the database and generate the training/ testing samples. The samples are sequences of voltage, current, charging capacity, charging energy, total charging capacity, total charging energy with a length of 120 s and frequency of 1 Hz, and their corresponding short circuit resistances. A big database with ∼ 6 × 105 samples are generated, covering various short circuit resistances (47 ∼ 470 Ω), current loading modes (Constant current-constant voltage (CC-CV) and drive cycle), and electrochemical states (cycle numbers from 1 to 300). Results show that the average relative absolute error of five random sample splits is 6.75 %±2.8 %. Further parametric analysis indicates the accuracy estimation benefits from the appropriate model setups: the optimized input sequence length (∼120 s), feature selection (at least one total capacity-related variable), and rational model design, using multiple layers with different kernel sizes. This work highlights the capabilities of machine learning algorithms and data-driven methodologies in real-time safety risk prediction for batteries.

The impact of the COVID-19 pandemic in diabetes and dyslipidemia management in a Spanish region: a retrospective study of the Aragon population.

Previous research has indicated that the COVID-19 outbreak had a negative impact on the diagnosis and management of cardiometabolic diseases. Our aim was to analyze the impact of the COVID-19 pandemic on the management of dyslipidemia and type 2 diabetes (T2D) in the Aragon region of Spain. We conducted an observational retrospective study, which included data from all patients diagnosed with active T2D or dyslipidemia in Aragon during 2019-2021. Data was collected from the BIGAN platform, a big database that includes all healthcare data from the Aragon population. Clinical, biochemical, and pharmacological prescription information was obtained for each patient and for each year. Out of the total population of 1,330,000 in the Aragon region, 90,000 subjects were diagnosed with T2D each year, resulting in a prevalence of approximately 7%. The COVID-19 pandemic resulted in a decrease in the prevalence of this disease and a lower incidence during the year 2020. In addition, patients with T2D experienced a deterioration of their glucose profile, which led to an increase in the number of patients requiring pharmacological therapy. The prevalence of dyslipidemia was approximately 23.5% in both 2019 and 2020 and increased to 24.5% in 2021. Despite the worsening of the anthropometric profile, the lipid profile improved significantly throughout 2020 and 2021 compared to 2019. Moreover, the number of active pharmacological prescriptions increased significantly in 2021. Our findings suggest that the overload of the health system caused by the COVID-19 pandemic has resulted in an underdiagnosis of T2D. Moreover, patients with T2D experienced a worsening of their glycemic profile, an increase in their pharmacological requirements, and lower performance of their analytical determinations. Dyslipidemic subjects improved their lipid profile although the value of lipid profile determination decreased between 2020 and 2021.

A Systematic Literature Mapping on Using Blockchain Technology in Identity Management

Although blockchain is an emerging technology, it has been applied in a lot of domains by leveraging its features. Traditional identity management systems have many issues regarding security and privacy of personal data. Blockchain has the potential to mitigate and avoid such issues by creating trust among the parties involved in the system while reducing reliance on third-party authorities. The first blockchain-based identity management solutions were launched in 2016. Since then, due to high demand, numerous primary and experimental studies and intatives have been carried out to provide solutions to this research topic. Along with that, there are also a lot of secondary studies to overview the current state of research on this topic. However, the number of systematic research articles is still limited and each research has it limitation. Through this study, we provide a novel systematic literature including categorization of studies into predefined categories (domain, research type, place of publication), analysis of publication frequency, co-authorship, number of papers citing each paper of all studied papers. Comparing to other systematic literature mapping studies, our paper provides a more comprehensive view of the studied articles. In particular, we analyze the number of citations, which no study has ever done. In this research, we studied 361 papers published from January 2009 to April 2022 in four big databases (IEEE Explore, ACM Digital Library, ScienceDirect, Springer Link), the largest number of articles studied compared to previous researches. The obtained results show that most of the articles under validation research type (providing solution and implementing that solution but not in real-world scenarios) propose solutions/systems, models/schemes and architectures to address general problems. We also find that the majority of authors works alone or collaborate in a separate group and co-work in only one paper. This shows that there is no long term collaboration in blockchain-based IdM identity management, and thus subsequent publications presenting real-world blockchain-based identity management products do not exist.

Power Prediction of VLSI Circuits Using Machine Learning

The difference between circuit design stage and time requirements has broadened with the increasing complexity of the circuit. A big database is needed to undertake important analytical work like statistical method, heat research, and IR-drop research that results in extended running times. This unit focuses on the assessment of test strength. Because of the enormous number of successful designs for current models and the unnecessary time required for every test, maximum energy ratings with all tests cannot be achieved. Nevertheless, test safety is important for producing trustworthy findings to avoid loss of output and harm to the chip. Generally, effective power assessment is only possible in a limited sample of pre-selected experiments. Thus, a key objective is to find the experiments that might give the worst situations again for testing power. It offers a machine-based circuit power estimation (ML-CPE) system for the selection of exams. Two distinct techniques of predicting are utilized. Firstly, to find testings with power dissipation, it forecasts the behavior of testing. Secondly, the change movement and energy data are linked to the semiconductor design, identifying small problem areas. Several types of algorithms are utilized. In particular, the methods compared. The findings show great accuracy and efficiency in forecasting. That enables such methods suitable for selecting the worst scenario.

Predicting smart cities’ electricity demands using k-means clustering algorithm in smart grid

This work aims to perform the unified management of various departments engaged in smart city construction by big data, establish a synthetic data collection and sharing system, and provide fast and convenient big data services for smart applications in various fields. A new electricity demand prediction model based on back propagation neural network (BPNN) is proposed for China?s electricity industry according to the smart city?s big data characteristics. This model integrates meteorological, geographic, demographic, corporate, and economic information to form a big intelligent database. Moreover, the K-means clustering algorithm mines and analyzes the data to optimize the power consumers? information. The BPNN model is used to extract features for prediction. Users with weak daily correlation obtained by the K-means clustering algorithm only input the historical load of adjacent moments into the BPNN model for prediction. Finally, the electricity market is evaluated by exploring the data correlation in-depth to verify the proposed model?s effectiveness. The results indicate that the K-mean algorithm can significantly improve the segmentation accuracy of power consumers, with a maximum accuracy of 85.25% and average accuracy of 83.72%. The electricity consumption of different regions is separated, and the electricity consumption is classified. The electricity demand prediction model can enhance prediction accuracy, with an average error rate of 3.27%. The model?s training significantly speeds up by adding the momentum factor, and the average error rate is 2.13%. Therefore, the electricity demand prediction model achieves high accuracy and training efficiency. The findings can provide a theoretical and practical foundation for electricity demand prediction, personalized marketing, and the development planning of the power industry.

Movie Recommender System Using Parameter Tuning of User and Movie Neighbourhood via Co-Clustering

Recommender System is a information filtering tool to filter the relevant information from given information in the present era of big data. Movie Recommender System is machine learning based autonomous tool that filters the movies from big movie database like netflix, amazon etc according to user preferences. The main focus of this paper is Partitional Weighted co-clustering for Movie Recommender System. The primary objective of this research article is to fine tune the parameters of user and movie neighborhoods by setting different values for row clusters number and column clutsers number parameters of co-clustering. Test results obtained from the Movie database show that the proposed method can bring more accurate personalized recommendations for the movie as compared to existing methods of the order of 7.91%.

Big Data Architecture Design of Smart Grid Based On Big Data Platform

A novel big data-driven approach has been proposed to address the challenges of extensive data and user application difficulties in the smart grid. A smart grid big data architecture based on a big data platform has been designed. The random matrix theory is applied to establish a random matrix algorithm model, which samples data from the smart grid big database and undergoes training and learning to construct a data model for user requirements. Through calculations, the macroscopic data hidden in the smart grid big data is transformed into microscopic data for users' reference, enabling the identification of parameters that affect the normal operation of the smart grid at its core. The computed data can be displayed locally and also uploaded remotely to the SG186 marketing system for various user applications. Additionally, the data can be permanently stored in the cloud through wireless communication. By incorporating the random matrix theory algorithm into the smart grid big data architecture, the proposed solution not only enhances the intuitive visualization of smart grid big data but also provides technical references for future work.

Fuzzy-HLSTM (Hierarchical Long Short-Term Memory) for Agricultural Based Information Mining

This research proposes a machine learning approach using fuzzy logic to build an information retrieval system for the next crop rotation. In case-based reasoning systems, case representation is critical, and thus, researchers have thoroughly investigated textual, attribute-value pair, and ontological representations. As big databases result in slow case retrieval, this research suggests a fast case retrieval strategy based on an associated representation, so that, cases are interrelated in both either similar or dissimilar cases. As soon as a new case is recorded, it is compared to prior data to find a relative match. The proposed method is worked on the number of cases and retrieval accuracy between the related case representation and conventional approaches. Hierarchical Long Short-Term Memory (HLSTM) is used to evaluate the efficiency, similarity of the models, and fuzzy rules are applied to predict the environmental condition and soil quality during a particular time of the year. Based on the results, the proposed approaches allows for rapid case retrieval with high accuracy.

FiRe 2 : An Online Database for Photographic and Cinematographic Film Technical Data

Among the great variety of our Cultural Heritage, photographic and cinematographic materials are fundamental and direct witnesses of the past. As often happens when dealing with materials of cultural interest, even photographic and cinematographic films can undergo severe deterioration, aging, and color fading throughout the time, leading to the need for proper restoration, conservation, and preservation of these materials. In this context, even though the digitization process is becoming more and more essential, analyzing and studying the analog materials remains fundamental to perform a retrieval or a correction faithful to the original material. For this reason, the lack of technical information (especially for the oldest materials) and the absence of open source archives of the producing companies underlines the actual and concrete need for a database of physical, chemical, and sensitometric data of both cinematography and photographic films. The aim of this work is the creation and the promotion of FiRe 2 , a big and unique database of cinematographic and photographic materials to support the work of conservators, restorers, and researchers, as the availability of information and the cooperation between institutions and professionals are essential for preserving our Cultural Heritage.

Clinical Outcomes Based on the Attainment of Low-Density Lipoprotein Cholesterol Targets in Patients with Acute Coronary Syndrome in Real-World Practice

Objective. A target of low-density lipoprotein cholesterol (LDL-C) <70 mg/dL or ≥50% reduction should be set. This study aimed to explore the information required to attain the optimal goal of lipid control for patients with ACS in real-world practice using big database analysis. Methods. Patients with ACS were enrolled between January 2005 and December 2019, and their medical history was obtained from the Chang Gung Research database. According to the attainment of LDL-C levels, the study population was divided into groups with and without ≥50% reduced LDL-C levels. In the group that achieved ≥50% reduced LDL-C levels, the study population was subdivided into groups with and without achievement of LDL-C level < 70 mg/dL. Results. This study enrolled 14,520 participants, out of whom only 3,367 patients (23.2%) achieved ≥50% reduced LDL-C levels. At the 3-year follow-up periods, higher incidences of cardiovascular (CV) mortality and all-cause mortality were absorbed in patients without ≥50% reduced LDL-C levels, especially in subgroups of hypertension and diabetes mellitus (DM). When comparing different percentages of reduced LDL-C levels, the significantly lowest hazard ratio (HR) of CV and all-cause mortality was noted at ≥50% reduced LDL-C levels (CV mortality; HR: 0.64; all-cause mortality; HR: 0.57). Conclusion. In the ACS population, better clinical outcomes were yielded in patients with ≥50% reduced LDL-C levels, especially in the hypertension and DM populations. However, strict lipid control did not show better clinical outcomes in patients with ≥50% reduction and <70 mg/dL in LDL-C levels.

Identifying Retinal Diseases on OCT Image Based on Deep Learning

Computer-aided diagnosis has the potential to replace or at least support medical personnel in their everyday responsibilities such as diagnosis, therapy, and surgery. In the area of ophthalmology, artificial intelligence approaches have been incorporated in the diagnosis of the most frequent ocular disorders, such as choroidal neovascularization (CNV), diabetic macular oedema (DMO), and DRUSEN; these illnesses pose a significant risk of vision loss. Optical coherence tomography (OCT) is an imaging technology used to diagnose the aforementioned eye disorders. It enables ophthalmologists to see the back of the eye and take various slices of the retina. The goal of this research is to automate the diagnosis of retinopathy, which includes CNV, DME, and DRUSEN. The approach employed is a deep learning-based, and transfer learning technique, applying to a public dataset of OCT pictures and two pertained neural network models VGG16 and InceptionV3, which are trained on the big database "ImageNet." That allows them to be able to extract the main features of millions of images. Furthermore, fine-tuning approaches are applied to outperform the feature extraction method, by modifying the hyperparameters. The findings showed that the VGG16 model performed better in classification than the InceptionV3 architecture, with a 0.93 accuracy.

Secretory autophagy promotes Rab37-mediated exocytosis of tissue inhibitor of metalloproteinase 1

BackgroundRab37-mediated exocytosis of tissue inhibitor of metalloproteinase 1 (TIMP1), an inflammatory cytokine, under serum-depleted conditions which leads to suppression of lung cancer cell metastasis has been reported. Starvation is also a stimulus of autophagic activity. Herein, we reveal that starvation activates Rab37 and induces autophagy.MethodsWe used an overexpression/knockdown system to determine the relationship between autophagy and Rab37 in vitro and in vivo. The autophagy activity was detected by immunoblotting, transmission electron microscope, autophagosome purification, and immunofluorescence under the confocal microscope. Lung-to-lung metastasis mouse model was used to clarify the role of autophagy and Rab37 in lung cancer. Clinical lung cancer patient specimens and an online big database were analyzed.ResultsInitially, we demonstrated that active-form Rab37 increased LC3-II protein level (the marker of autophagosome) and TIMP1 secretion. Accordingly, silencing of Rab37 gene expression alleviated Rab37 and LC3-II levels as well as TIMP1 secretion, and induction of autophagy could not increase TIMP1 exocytosis under such conditions. Moreover, silencing the Atg5 or Atg7 gene of lung cancer cells harboring active-mutant Rab37 (Q89L) led to decreased autophagy activity and TIMP1 secretion. In the lung-to-lung metastasis mouse model, increased TIMP1 expression accompanied by amiodarone-induced autophagy led to decreased tumor nodules and cancer cell metastasis. These phenomena were reversed by silencing the Atg5 or Atg7 gene. Notably, increasing autophagy activity alone showed no effect on TIMP1 secretion under either Rab37 or Sec22b silencing conditions. We further detected colocalization of LC3 with either Rab37 or TIMP1, identified Rab37 and Sec22b proteins in the purified autophagosomes of the lung cancer cells harboring the active-form Rab37 gene, and confirmed that these proteins are involved in the secretion of TIMP1. We reveal that autophagic activity was significantly lower in the tumors compared to the non-tumor parts and was associated with the overall lung cancer patient survival rate.ConclusionsWe are the first to report that autophagy plays a promoting role in TIMP1 secretion and metastasis in a Rab37-dependent manner in lung cancer cells and the lung-to-lung mouse model.

Confidence, community & conservation: Exploring the relationship between self-efficacy and experience in female hunters

Self-efficacy, or the belief we have in our own abilities, plays an important role in determining individuals’ participation and success in outdoor activities. In the U.S., managers have increasingly sought ways to support underrepresented groups’ participation in outdoor recreation. Hunting provides a particularly illustrative example, where female hunters and hunters of other racial and ethnic groups remain a starkly underrepresented, albeit growing constituent of the U.S. hunting population. In this study, we investigated the role of self-efficacy as it relates to female hunter participation to inform managers’ innovative efforts to recruit and retain this important constituency. Specifically, we look at how self-efficacy and its components change as female hunters gain experience. In a sample of female Oregon hunters (n = 147) drawn from the 2008 big game license database conducted in the summer of 2010, we found that hunters with fewer years of experience had lower overall self-efficacy compared to more experienced hunters. While skills-based components of self-efficacy were lower for less experienced hunters, there was less of a difference in the social support-based components of self-efficacy as hunters gained experience. These findings suggest that social support is important for the recruitment and retention of all female hunters regardless of skill level or experience. Managers seeking to bolster or maintain hunter participation might consider tailoring recruitment and retention efforts to address the social support needs and unique motivations of female hunters as they seek to achieve the goals of inclusivity in hunting as well as conservation and wildlife management more broadly.

A review of civil society organizations concepts through content analysis

Day by day, civil society organizations (CSOs) are becoming an increasingly important factor in all areas of the socio-economic system. The interest in their study has also increased, although the place, role, and definition of their terminology and concepts, needs to be further clarified, as their meaning changes conditionally. Through content analysis, this article reviews the literature by addressing a sample of 136 journal articles published in the period 2015-2019 in two big databases. The findings show that CSOs literature is dominated by qualitative works of social sciences authors who mostly use the term Non-Governmental Organizations (NGOs). This paper is also elaborated on the main themes which dominate the text of the articles. Furthermore, the main concepts of this literature, development, organizations, international, local and other words are analyzed here. The significance of this study is that it aims to provide CSOs researchers with a clearer description of the terminology and concepts the most commonly used.

Prediction of attachment efficiency using machine learning on a comprehensive database and its validation

Colloidal particles can attach to surfaces during transport, but the attachment depends on particle size, hydrodynamics, solid and water chemistry, and particulate matter. The attachment is quantified in filtration theory by measuring attachment or sticking efficiency (Alpha). A comprehensive Alpha database (2538 records) was built from experiments in the literature and used to develop a machine learning (ML) model to predict Alpha. The training (r-squared: 0.86) was performed using two random forests capable of handling missing data. A holdout dataset was used to validate the training (r-squared: 0.98), and the variable importance was explored for training and validation. Finally, an additional validation dataset was built from quartz crystal microbalance experiments using surface-modified polystyrene, poly (methyl methacrylate), and polyethylene. The experiments were performed in the absence or presence of humic acid. Full database regression (r-squared: 0.90) predicted Alpha for the additional validation with an r-squared of 0.23. Nevertheless, when the original database and the additional validation dataset were combined into a new database, both the training (r-squared: 0.95) and validation (r-squared: 0.70) increased. The developed ML model provides a data-driven prediction of Alpha over a big database and evaluates the significance of 22 input variables.

Factors affecting urban climate at different times of the day in China: A case study in Yibin, a riverside mountain city

Urban heat islands and global warming degrade urban climate substantially and negatively affect residents’ lifestyle. Experts have explored strategies to mitigate their effects. They found that vegetation and water surfaces had a significant cooling effect, which varied with the time of the day. Thermal performances of the two elements echoed the strategies of nature-based solutions in cooling. Meteorology stations in different counties of one city are sources of big databases providing regionally long-term meteorological data, available for climate studies. The present study used this data to investigate the variation in cooling effects in Yibin, a riverside mountain city located in China. The cooling effects of vegetation and water surface (indicated by river width) were significantly affected by the time of the day. Increasing vegetation cover reduced air temperature by a maximum of 2.2 °C (R² = 0.615, at 4:00 h). A similar cooling effect was observed with river (R² = 0.5149, at 14:00 h). Nevertheless, the cooling effects fluctuated during the day. Vegetation had significant thermal effects in the late afternoon and early evening (4:00–7:00), which is the coolest time during the day. Conversely, waterbodies showed limited effect on weather conditions in the evening. They were thermally adjustable in the afternoon (13:00–16:00, daily warmest time). This phenomenon is attributed to hourly weather changes. Overall, R² values of vegetation-temperature models decreased, while those of water-temperature models increased with temperature. These findings represent an advance in our understanding of urban climate. Also, it is important to find that the two key cooling factors (vegetation and river width) behave differently at different times of the day. To achieve full day cooling, it is imperative that future planning should incorporate proper landscaping in their design works. This finding is helpful for future generations in terms of climate change adaptation.

Using a Clustering Method to Detect Spatial Events in a Smartphone-Based Crowd-Sourced Database for Environmental Noise Assessment.

Noise has become a very notable source of pollution with major impacts on health, especially in urban areas. To reduce these impacts, proper evaluation of noise is very important, for example by using noise mapping tools. The Noise-Planet project seeks to develop such tools in an open science platform, with a key open-source smartphone tool "NoiseCapture" that allows users to measure and share the noise environment as an alternative to classical methods, such as simulation tools and noise observatories, which have limitations. As an alternative solution, smartphones can be used to create a low-cost network of sensors to collect the necessary data to generate a noise map. Nevertheless, this data may suffer from problems, such as a lack of calibration or a bad location, which lowers its quality. Therefore, quality control is very crucial to enhance the data analysis and the relevance of the noise maps. Most quality control methods require a reference database to train the models. In the context of NC, this reference data can be produced during specifically organized events (NC party), during which contributors are specifically trained to collect measurements. Nevertheless, these data are not sufficient in number to create a big enough reference database, and it is still necessary to complete them. Other communities around the world use NC, and one may want to integrate the data they collected into the learning database. In order to achieve this, one must detect these data within the mass of available data. As these events are generally characterized by a higher density of measurements in space and time, in this paper we propose to apply a classical clustering method, called DBSCAN, to identify them in the NC database. We first tested this method on the existing NC party, then applied it on a global scale. Depending on the DBSCAN parameters, many clusters are thus detected, with different typologies.